Professional Documents
Culture Documents
Annemieke van Straten, Ph.D., Tanja van der Zweerde, M.Sc., Annet Kleiboer, Ph.D.,
Pim Cuijpers, Ph.D., Charles M. Morin, Ph.D., Jaap Lancee, Ph.D.
PII: S1087-0792(17)30034-5
DOI: 10.1016/j.smrv.2017.02.001
Reference: YSMRV 1017
Please cite this article as: van Straten A, van der Zweerde T, Kleiboer A, Cuijpers P, Morin CM, Lancee
J, Cognitive and behavioral therapies in the treatment of insomnia: a meta-analysis, Sleep Medicine
Reviews (2017), doi: 10.1016/j.smrv.2017.02.001.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
PT
Authors:
Annemieke van Straten, Ph.D.1
RI
Tanja van der Zweerde, M.Sc.1
Annet Kleiboer, Ph.D.1
Pim Cuijpers, Ph.D.1
SC
Charles M. Morin, Ph.D.2
Jaap Lancee, Ph.D.3
U
Affiliations
1
Department of Clinical Psychology & EMGO Institute for Health and Care Research, VU University
Amsterdam, The Netherlands
2
AN
Université Laval, École de Psychologie, Québec City, QC, Canada
3
Department of Clinical Psychology, University of Amsterdam, The Netherlands
M
Corresponding author
D
1081 BT Amsterdam
The Netherlands
Phone: + 31205988970
C
Mobile: +31627230697
e-mail: a.van.straten@vu.nl
AC
Acknowledgements
Prof. Van Straten, Ms van der Zweerde, dr. Kleiboer, prof. Cuijpers and dr. Lancee, declare that they
have no competing interests. Prof. Morin has served as a consultant for Merck, Valeant, and Novartis
and received research support from Novartis. We did not receive any financial support to write this
meta-analysis.
1
ACCEPTED MANUSCRIPT
Summary
Insomnia is a major public health problem considering its high prevalence, impact on daily life, co-
morbidity with other disorders and societal costs. Cognitive behavioral treatment for insomnia (CBTI) is
currently considered to be the preferred treatment. However, no meta-analysis exists of all studies using
PT
at least one component of CBTI for insomnia, which also uses modern techniques to pool data and to
RI
analyze subgroups of patients. We included 87 randomized controlled trials, comparing 118 treatments
(3724 patients) to non-treated controls (2579 patients). Overall, the interventions had significant effects
SC
on: insomnia severity index (ISI; g=0.98), sleep efficiency (SE; g=0.71), Pittsburgh sleep quality index
(PSQI; g=0.65), wake after sleep onset (WASO; g=0.63) and sleep onset latency (SOL; g=0.57), number of
U
awakenings (NWAK; g = 0.29) and sleep quality (SQ; g = 0.40). The smallest effect was on total sleep
AN
time (TST; g=0.16). Face-to-face treatments of at least 4 sessions seem to be more effective than self-
help interventions or face-to-face interventions with fewer sessions. Otherwise the results seem to be
M
quite robust (similar for patients with or without comorbid disease, younger or older patients, using or
D
not using sleep medication). We conclude that CBTI, either its components or the full package, is
Keywords
Behavior therapy; cognitive therapy; cognitive behavior therapy; insomnia; sleep initiation or
C
maintenance disorder
AC
2
ACCEPTED MANUSCRIPT
List of abbreviations
PT
CMI = comorbid insomnia
CT = cognitive therapy
DSM = Diagnostic and statistical manual of mental disorders
RI
DSISD = Duke structured interview for sleep disorders
ICSD: International classification of sleep disorders
ISI: insomnia severity index
F2F = face-to-face
SC
NWAK = number of awakenings
NR = not reported
PE = psycho-education
U
PI = primary insomnia
PR = progressive relaxation
PSQI: Pittsburgh sleep quality index
RDC = research diagnostic criteria
RCTs = randomized controlled trials
AN
SC = stimulus control
M
SD = standard deviation
SE = sleep efficiency
SH = sleep hygiene
D
3
ACCEPTED MANUSCRIPT
Introduction
Insomnia is a major public health problem. The prevalence of insomnia, which may be characterized by
difficulty initiating or maintaining sleep, with significant distress and impairments of daytime
functioning, is high: about a third of the population suffers from insomnia symptoms, and about 10%
PT
5 fulfills the criteria for a sleep disorder [1, 2]. Insomnia has a high burden of disease, which impacts daily
RI
life in different domains [3, 4] and often persists for many years [5]. The societal costs are substantial:
poor sleepers cost society about ten times as much as good sleepers [6]. These societal costs are due to
SC
increased health care consumption but especially caused by reduced work productivity and increased
U
10 In addition, insomnia is strongly associated with other somatic and mental health problems as
AN
well as with an increased mortality rate [8-10]. Most notable is the association with cardiovascular
diseases [11-13] and with depression [14-16]. The nature of these associations is still not clear but it has
M
been suggested that hyperarousal and the chronic activation of stress responses is a possible pathway
D
between insomnia, depression and cardiac disease [13]. There are also indications that insomnia is a
TE
Treatment of insomnia is highly desirable, mostly to decrease the burden of insomnia itself. But
EP
it might also contribute to a decrease of the associated somatic and mental health problems such as
20 enhancing sleep in the short run, but with risks of negative side effects and limited evidence for their
sleep hygiene), behavioral (relaxation, sleep restriction, stimulus control, paradoxical intention) or
cognitive (identifying and challenging dysfunctional thoughts and excessive worries about sleep; [21-
4
ACCEPTED MANUSCRIPT
26]). Since the 1990s it has become popular to offer these non-pharmacological treatments in (various)
combinations. These combinations are usually referred to as cognitive behavioral therapy for insomnia
(CBTI).
Several excellent (systematic) reviews and meta-analyses have been written on CBTI; they
PT
5 conclude that CBTI is effective [27-33] in primary [34] and comorbid [35, 36] insomnia and CBTI is at
RI
least as effective as pharmacotherapy [37, 38]. As a result, the American College of Physicians recently
recommended CBTI as the initial treatment for all adults with insomnia [39]. Even though some of these
SC
reviews pooled the data of individual studies [e.g. 28, 29, 35], to our knowledge, no recent meta-analysis
exists that includes all CBTI studies and that uses modern techniques to pool data and to analyze the
U
10 subgroups of patients which might benefit most from CBTI.
AN
The aim of this meta-analysis is to quantify the effects of educational, behavioral and cognitive
therapies for insomnia, based on all available randomized controlled trials, and to perform subgroup
M
analysis as a function of several potential moderators (e.g. comorbidity, sleep medication, year of
D
15
TE
Method
EP
Search strategy
We carried out a comprehensive literature search in PubMed, PsycINFO, EMBASE and the Cochrane
C
central register of controlled trials. We combined terms indicative of insomnia (e.g. insomnia, sleep
AC
20 disorders, sleep initiation and maintenance disorders) with those of psychological treatment (e.g.
psychotherapy, cognitive therapy, behavior therapy). For example for PsycINFO we used (DE=("sleep
therapy" or "cognitive therapy")). We searched all literature up to December 2015. Titles and abstracts
were screened by one person (AvS, TvdZ, AK or JL). Only those records that were definitely not suitable
5
ACCEPTED MANUSCRIPT
(e.g. not a randomized trial, a biological or medical treatment) were excluded in this phase. We
retrieved the full papers of the remaining 196 references. Those papers were examined independently
by two of the four researchers (AvS, TvdZ, AK or JL). In case of disagreement the paper was discussed
with the third and fourth reviewer until consensus was achieved.
PT
5
RI
Inclusion criteria
We used the following inclusion criteria: 1) the study had to be a randomized controlled trial (RCT), 2)
SC
investigating CBTI or at least one component of it, 3) for adults of 18 years or older, 4) with insomnia, 5)
in comparison with a non-active control group (e.g. waitlist control, care-as-usual, or a minimal
U
10 intervention such as psycho-education about sleep or sleep hygiene information), 6) reporting on sleep
AN
diary outcomes. We identified the following treatments as being part of CBTI: psycho-education, sleep
hygiene, relaxation, sleep restriction, stimulus control, paradoxical intention, and identifying and
M
challenging dysfunctional thoughts (about sleep). We excluded all other therapies such as interpersonal
D
therapy, bright light therapy, exercise and biofeedback. Cognitive distraction was also excluded because
15 it only encompassed advice on topics to think about (e.g. a recipe or a plot of a television show) or
TE
things to do (e.g. to read a book or watch television). We also excluded studies: aimed at children or
EP
adolescents, at tapering of medication, which used outcomes such as fatigue instead of sleep, or which
were aimed at treating another mental health disorder and reported insomnia as a secondary outcome.
C
We also excluded studies with insufficient information to calculate the effect size (e.g. those presenting
AC
20 scores in a plot without providing means, standard deviations or other relevant statistics).
Data extraction
We coded the following characteristics of the studies: 1) year of publication, 2) setting where patients
were recruited (community, primary care, other care facilities, university), 3) the definition of insomnia
6
ACCEPTED MANUSCRIPT
that was used, 4) co-morbidity (e.g. insomnia in chronic pain patients), 5) age group (young adults, older
adults or all adults), 6) the treatment format (individual, group or self-help), 7) the number of treatment
sessions, 8) the use of sleep medication (allowed, not allowed or not reported), 9) the type of control
group (e.g. waitlist, no treatment), 10) the number of patients included in the treatment and control
PT
5 group, and 11) type of intervention. We distinguished four categories of interventions: a) full CBTI which
RI
had to include an educational component as well as a behavioral and cognitive one, b) behavioral
therapy which had to include both stimulus control and sleep restriction (with or without an educational
SC
component), c) relaxation, and d) “other” which included e.g. stimulus control only or paradoxical
intention only. Two independent assessors coded each study and differences were discussed by the
U
10 review team until consensus was reached (AvS, TvdZ, AK, or JL).
AN
Quality assessment
M
We assessed the validity of the studies using the criteria suggested by the Cochrane Handbook [40]: 1)
D
15 outcome data, and 4) selective reporting of data. We did not assess the blinding of patients or therapists
TE
since this is impossible in psychotherapy research nor did we assess blinding of outcome assessors since
EP
all reported outcomes are based on self-report. Two reviewers conducted the quality assessment
20 Meta-analyses
We examined the effects as observed immediately after treatment (post-test). We calculated between
group effect sizes for all individual studies using Hedges’ g. This is the standardized mean differences (or
Cohen’s d) after adjusting for small sample sizes [41, 42]. The effect size represents the difference
between two groups in number of standard deviations. Effect sizes of 0.56-1.2 can be assumed to be
7
ACCEPTED MANUSCRIPT
large, while effect sizes of 0.33-0.55 are moderate, and effect sizes of 0-0.32 are small [43]. To calculate
those effect sizes, we used the available statistics as published in the papers (e.g. means and standard
To calculate the individual effect sizes as well as the pooled mean effect size we used the
PT
5 computer program Comprehensive meta-analysis (CMA) version 3.3.070 for Windows, developed for
RI
support in meta-analysis (www.metaanalysis.com). We first performed the analyses on all studies and
then checked for outliers. An outlier was defined as a study in which the 95% confidence interval around
SC
the effect size did not overlap with the 95% confidence interval around the pooled effect size. We then
U
10 As we expected considerable heterogeneity, we calculated pooled effect sizes with the random
AN
effects model. We tested heterogeneity under the fixed effects model using I2 which describes the
variance between studies as a proportion of the total variance. A value of 0% indicates no observed
M
heterogeneity and larger values show increasing heterogeneity. We calculated 95% confidence intervals
(CI) around I2, using the non-central Chi squared-based approach within the heterogi module for Stata.
D
15 We tested for publication bias by visually inspecting the funnel plot and by conducting the
TE
Egger’s test of the intercept. We used the the Duval and Tweedie [44] trim and fill procedure to estimate
EP
the effect size after the publication bias had been taken into account. It also provides an estimate of the
20 sleep onset latency (SOL), sleep efficiency (SE) and insomnia severity index (ISI). We chose SOL because
this was the most frequently used outcome variable, SE because this is often described as the preferred
primary outcome in insomnia research and ISI because this questionnaire is gaining popularity rapidly as
an outcome and is much easier to collect than a sleep diary. For these three variables we tested whether
the effect size was significantly related to treatment, patient or study variables. We used the mixed
8
ACCEPTED MANUSCRIPT
effects model, which pools studies within subgroups with the random effects model, but tests for
significant differences between subgroups with the fixed effects model. Lastly, we conducted
multivariate meta regression analysis. We only included variables that were univariate associated with
the outcome with a p-value of 0.25 or less. Two variables were excluded because of collinearity: type of
PT
5 treatment (which was collinear with yes/no full CBTI treatment) and number of sessions (which was
RI
collinear with treatment format since self-help interventions cannot be divided into discrete sessions).
We entered all the variables in the model simultaneously and calculated standard regression
SC
coefficients. Finally, we reported the overall proportion of the total between-study variance which is
U
10
Results
AN
Selection of studies
M
Figure 1
D
The titles and abstracts of 1727 references were screened (after removal of 290 duplicates). We
TE
15 excluded 1503 references and retrieved full-text papers of the remaining 224 references. A total of 137
papers did not fulfill our inclusion criteria (Figure 1). We included 87 papers on RCTs in which (a
EP
component of) CBTI was examined in comparison to a non-treatment control group [45-131]. Some
studies had more than two arms and examined different active interventions in comparison to a control.
C
Of the 87 studies, 63 examined one intervention, 18 examined two interventions, 5 studied three
AC
20 interventions and 1 studied four interventions. The total number of comparisons was therefore 118.
There were 6.303 patients included in the studies, 3.724 in the intervention groups and 2.579 in the
control groups.
9
ACCEPTED MANUSCRIPT
Table 1
The first studies were published in 1974 but the majority of studies (n=62; 71%) were published in or
after the year 2000 (Table 1). Most studies recruited people from the general population (n=59; 68%). A
minority of the studies recruited people through care facilities (primary care or specific care settings,
PT
5 n=20; 23%) and many of those studies additionally recruited via general media. The remaining 8 studies
RI
(9%) recruited participants through universities. Some studies excluded older patients (n=18; 21%),
while others specifically focused on the elderly (n=20; 23%). The remaining studies either included both
SC
younger and older adults (n=30; 34%) or did not specify which age categories were eligible for the study
(n=19; 22%). Some studies specifically examined insomnia in the context of another somatic disorder
U
10 (n=15; 17%) or a mental disorder (n=5; 6%) but most studies aimed to examine primary insomnia and
AN
therefore excluded patients with mental or somatic illnesses (n=45; 52%). For the remaining studies
comorbidities were allowed but it was not always clear whether or not they excluded patients with
M
severe illnesses (n=22; 25%). About half of the studies allowed the use of sleep medication during the
D
study (n=40; 46%), while in the other half of the studies sleep medication was either not allowed (n=37;
15 43%) or it was not reported how the use of sleep medication was handled (n=10; 11%).
TE
Out of the 118 interventions included in this meta-analysis 51 (43%) consisted of full CBTI
EP
(educational component as well as a behavioral and cognitive one), 13 (11%) of behavioral therapy
(stimulus control and sleep restriction but no cognitive element), 23 (19%) of relaxation (without any
C
other treatment element) while the remaining 31 interventions (26%) consisted of “other” treatments
AC
20 such as paradoxical intention only, or stimulus control only. About a third of the treatments (n=35; 30%)
was offered in group format, about half (n=57; 48%) was offered as individual face-to-face therapy while
the remaining studies examined self-help treatments (either through books, audio or the Internet, n=26;
22%). The individual or group treatments typically took up to 6 sessions (n=71; 60%). Of the 87 studies,
10
ACCEPTED MANUSCRIPT
41 used a waitlist control (47%), 14 no treatment (16%), 20 placebo (mostly quasi-desensitization; 23%)
Quality assessments
PT
5 Out of all 87 studies 38 (44%) reported an adequate sequence generation for their randomization while
RI
the remaining 49 studies did not report how they handled this. Only 23 studies (26%) reported that their
random allocation had been concealed. For the other 64 studies (74%) this remained unclear. About half
SC
of the studies (n=45; 52%) handled their missing data well by performing intention-to-treat analyses. For
another 24 studies (27%) it was clear that missing cases had been dropped from their analysis. For the
U
10 other 18 studies (21%) it remained unclear how many patients were randomized, how many patients
AN
had dropped out and how missing data was handled.
We used self-report sleep diary data to extract our outcome data. Many different variables
M
might be extracted from a sleep diary and this was reflected in the included studies. Of all the 118
D
comparisons 108 calculated SOL. This is the time it takes to fall asleep after going to bed and turning the
15 lights out/attempting to sleep. The SE, the percentage of time having slept while being in bed, was
TE
calculated for 79 comparisons, total sleep time (TST) for 91, the time being awake after sleep onset
EP
(WASO) for 71, sleep quality (SQ) for 40 and the number of awakenings during the night (NWAK) for 36
comparisons. Out of all 87 studies 38 also included the insomnia severity index (ISI [132]) as an outcome
C
measure and 19 included the Pittsburg sleep quality index (PSQI [133]).
AC
20
Overall effects
Table 2
The CBTI interventions had positive statistically significant effects on all reported sleep outcomes (Table
2). The ISI showed the largest effect (g=0.98) but large effects were also obtained for SE (g=0.71), PSQI
11
ACCEPTED MANUSCRIPT
(g=0.65), WASO (g=0.63) and SOL (g=0.57). Small to moderate effect sizes were observed for NWAK (g =
0.29) and SQ (g = 0.40). The smallest effect was on TST (g=0.16). Since there were studies with
extremely high effect sizes as well as studies with extremely low effect sizes, deleting the outliers did
not influence the effect sizes substantially. It did reduce the percentage of heterogeneity (I2).
PT
5 Heterogeneity remained quite high for the ISI and the SQ but was small to moderate for all outcomes
RI
based on sleep diaries. Of the 87 studies, 24 (28%) examined more than one treatment while using only
one control group for each study. This means that the comparisons from these studies were not
SC
independent of each other which may have resulted in an artificial reduction of heterogeneity. We
examined the possible effects by conducting sensitivity analyses in which we included only one
U
10 comparison per study. First, we included only the comparison with the largest effect size from that study
AN
and then we conducted another analysis in which we included only the smallest effect size. The resulting
effect sizes and heterogeneity statistics were highly comparable with the ones found in the overall
M
analyses (Table 2) and therefore not shown.
D
15 Table 3
TE
We studied the association between the effect size on SOL, SE and ISI on the one hand and different
EP
treatment, patient and study variables on the other (Table 3). Overall the results were mixed and
First, we describe the results for SOL. SOL was significant (p < 0.05) related to type of treatment,
AC
20 age, and year of publication: effect sizes were higher for relaxation and “other” treatments than for CBTI
or behavioral treatments and higher for young adults than for older adults. There seems to be a U-shape
association between SOL and the year of publication: the oldest studies have the highest effect sizes,
but after the 90s the effect sizes start to increase again. There were also significant associations for the
use of sleep medication and concealment of allocation. In these cases effect sizes were higher for
12
ACCEPTED MANUSCRIPT
studies that did not report the use of sleep medication and for studies that did not report allocation
concealment. There were borderline significant (p < 0.10) associations for SOL with treatment format
(individual and group treatments yielded higher effect sizes than self-help), number of sessions (longer
treatments yielded better effect sizes than shorter treatments), recruitment population (non-
PT
5 community samples yielded higher effect sizes), and type of control (no treatment had highest effect
RI
size). Overall, heterogeneity seemed to be moderate.
Second, we describe the results for SE. Some of the results for SE were comparable to those of
SC
SOL. This means that there was a (borderline) significant association with: type of treatment (higher
effect sizes for “other” treatment than for CBTI or behavioral treatments), treatment format (higher
U
10 effect sizes for individual and group therapy than for self-help), number of sessions (higher effect sizes
AN
for 5 or more sessions than for 1 to 4), age (higher effects for younger people than for older people),
recruitment population (higher effect sizes for patients recruited in other ways than general media /
M
community) and year of publication (the most recent publications yielded the highest effect sizes). There
D
were also some differences between the results for SOL and SE because SE was also (borderline)
15 significantly associated with comorbidity (people with comorbid disorders obtained higher effects) and
TE
ITT analysis (studies using ITT analysis showed higher effects than studies which did not report this).
EP
Third, we describe the results for ISI. ISI was significantly related only to the type of control:
studies with waitlist controls showed higher effects than studies with other type of control groups.
C
There was furthermore a borderline significant association with allocation concealment: studies not
AC
Table 4
We examined a number of variables which might be related to each other (e.g. studies
examining self-help might have included younger people than studies examining individual therapy).
Therefore, we also performed a multivariate analysis. The multivariate model showed significant
13
ACCEPTED MANUSCRIPT
relations with SE and ISI for treatment format. Self-help interventions performed worse than individual
face-to-face interventions. Two other significant associations were related to study quality: studies that
did not report allocation concealment had higher effects on SOL and ISI, and studies that did not report
age had higher effects on SE and ISI. Finally, two significant associations were related to study design:
PT
5 studies with a waitlist control showed higher effect sizes on ISI than studies with an information control
RI
(or other minimal interventions) and more recent studies showed higher effect sizes on SE than older
studies.
SC
Publication bias
U
10 Visual inspection of the funnel plot seemed to indicate that there was publication bias and this was
AN
confirmed by Egger’s test (p < 0.01 for SOL, SE and ISI). This means that there are studies which are not
published because of their small effects. Duval and Tweedie have developed a method (implemented in
M
CMA) which is able to estimate how many studies are missing and what their effect size would have
D
been. It re-calculates the pooled effect size while imputing the missing effect sizes. The results showed
15 that 27 studies were missing for SOL, 21 for SE, and 13 for ISI. The effects remained statistically
TE
significant after re-calculation but became smaller (g=0.41 for SOL, g = 0.49 for SE, and g = 0.71 for ISI;
EP
Table 3).
C
Discussion
AC
20 In this meta-analysis on 87 studies with 118 comparisons we examined the effects of educational,
behavioral and / or cognitive treatments for insomnia. The overall effects were large on insomnia
severity (ISI; g = 0.98), sleep efficiency (SE; g=0.71), the Pittsburgh sleep quality Index (PSQI; g=0.65),
wake after sleep onset (WASO; g=0.63) and sleep onset latency (SOL; g=0.57). Small to moderate effect
sizes were observed for number of awakenings (NWAK; g = 0.28) and sleep quality (SQ; g = 0.40). The
14
ACCEPTED MANUSCRIPT
smallest effect was on total sleep time (TST; g=0.16). Heterogeneity was highest for self-reported
outcomes (ISI and SQ) and more moderate for sleep diary variables. Multivariate analysis showed
different associations for different sleep variables (SOL, SE and ISI). We found that face-to-face
interventions perform better than self-help interventions (on SE and ISI but not on SOL), newer studies
PT
5 and studies with a waitlist control yielded higher effect sizes (on SE and ISI respectively), and studies
RI
that did not report age or allocation concealment reported higher effect sizes (SOL and ISI, and SE and
ISI respectively).
SC
The overall effects for insomnia treatments seem to be in line with psychological treatments for
other disorders such as depression (d = 0.78 [134]), social phobia (d=0.70 [135]), panic disorder with
U
10 (d=0.78) or without agoraphobia (d=1.04 [136]). However, the magnitude of the effects varied between
AN
the different outcome parameters. Quite large effect sizes were obtained for scores on the
questionnaires of insomnia severity (ISI) and sleep quality (PSQI) even though there was considerable
M
heterogeneity. The ISI and PSQI do not rate the nocturnal symptoms of insomnia alone but also their
D
daytime consequences. However, it would be premature to conclude from this that insomnia
15 treatments are effective in improving daytime functioning. We need studies examining day-time
TE
functioning in more detail (e.g. examining fatigue, work productivity, social activities) before we can
EP
Several indices derived from the daily sleep diary (SE, WASO and SOL) also yielded high effect
C
sizes, indicating that treatment was very effective for improving sleep continuity. As previously reported
AC
20 in other studies, the actual number of hours of sleep (TST) improved only modestly. Post-hoc analysis
did not seem to indicate that TST was related to type of treatment indicating that TST was not
(artificially) reduced because of sleep restriction. Thus, the findings indicate that treatment is mostly
effective for improving sleep continuity/efficiency and sleep quality, with some evidence of improved
daytime functioning as well. It must be noted, however, that not all outcomes were reported in all
15
ACCEPTED MANUSCRIPT
studies. This hampers the interpretation of the differences in effect sizes on the various measures. Our
knowledge about insomnia and its treatments would be greatly enhanced if future randomized clinical
trials would use the same outcome measures to estimate treatment effects [137].
Overall, CBTI treatments for insomnia are effective but the next question is: can we explain the
PT
5 differences in effects between studies (the heterogeneity)? First we examined this by comparing the
RI
studies that used full CBTI to those who did not include all CBTI components. There were no statistically
significant differences between the effects of those studies. Second we compared full CBTI to different
SC
(combinations of) components. Univariate analyses showed that relaxation works well for SOL and that
“other” treatments work the best on all sleep outcomes. The high effects for “other” treatments
U
10 compared with full CBTI are counterintuitive. One explanation might be that these studies tend to be
AN
older and these older studies tend to have poorer quality (no report on randomization sequence or
randomization concealment). The effect sizes for the full CBTI package seem to be similar to those for
M
behavioral techniques only. This brings into question the additional effect of the cognitive modules. One
D
recent study was aimed at this question and compared behavioral treatment (BT) with cognitive
15 treatment (CT) and full CBTI [138]. Even though this study did not demonstrate significant differences
TE
among the three treatments on the post-test continuous ISI scores, there were some indications that BT
EP
worked faster than CT but that the effects of CT lasted longer than those of BT. Therefore, to make use
of both effects, the authors recommended full CBTI. Our current meta-analysis, which did not include
C
20 Another reason for the failure to detect clear and consistent differences in effects between the
different treatments is that there is considerable variety in the treatments within one category (e.g. full
CBTI). This has been described previously with respect to sleep restriction treatment [139]: there was a
great variability in the way the average number of hours asleep was calculated, the minimum number of
hours in bed, how bed and rise times were positioned, what the sleep efficiency criterion was for
16
ACCEPTED MANUSCRIPT
extending the sleep window, for how many days this criterion should be reached before actually
extending etc. In all likelihood, these variations also exist for the other CBTI elements. Our data support
this idea since we demonstrated quite large 95% confidence intervals for heterogeneity within the
different treatment subgroups. Ultimately we need to standardize our treatment components and
PT
5 examine differences in effect by direct comparisons. We recommend that more dismantling studies be
RI
performed.
The treatment format was significantly related to all studied sleep outcomes. More specifically:
SC
self-help interventions seem to perform worse than face-to-face interventions, especially on SE and ISI.
This is in contrast to studies on other mental disorders which often yield a similar or quite similar effect
U
10 size for self-help interventions compared with face-to-face interventions [140, 141]. Moreover, some of
AN
the recent studies on self-help interventions demonstrated very high effect sizes, even those without
any human support [64, 98]. Since insomnia is so prevalent and self-help interventions are much easier
M
to offer than face-to-face treatments against lower costs because of reduced therapeutic input, we
D
conclude that self-help interventions are worthwhile as a first step of treatment. But sufficient face-to-
15 face options need to be available for those who do not recover. Furthermore, we urgently need studies
TE
comparing the two treatment formats directly since two recent studies showed mixed results with one
EP
not finding differences in effect [142] and the other finding face-to-face treatment superior to online
treatment [87].
C
It remains unclear to what extent the number of treatment sessions is important to obtain
AC
20 optimal effects. There are some indications from the univariate analysis that 1 to 4 session treatments
yield lower effect sizes than treatments which are composed of more than 4 sessions. However, these
results are not very robust and do not apply to the ISI. It was not possible to test the effect of the
number of session in a multivariate analysis since the self-help interventions all had 0 sessions and this
caused collinearity. The only study that directly compared the treatment duration (1, 2 ,4 and 8 sessions;
17
ACCEPTED MANUSCRIPT
[61]) concluded that a duration of 4 sessions was optimal. Since Edinger and colleagues [61] concluded
that 4 sessions were optimal, and we have indications that 5 or more sessions are more effective, we
conclude that 1 to 3 session treatments seem to be less effective than longer treatments. However,
more studies directly comparing treatment length (and costs) are needed to establish optimal treatment
PT
5 length.
RI
Interestingly, there seems to be an association between publication year and treatment effect.
For SE the newest studies (after 2010) showed significantly higher effect sizes than those published
SC
between 1990 and 2010. There were no studies measuring SE before 1990. For SOL, there seemed to be
a U-shaped association: the oldest studies showed the highest effects. Effect sizes then decreased over
U
10 time but the newest studies yielded higher effects again. The association disappeared in the multivariate
AN
analysis but this might be because in this analysis we used the actual publication year instead of
categories. Furthermore, we tested for a linear relationship and not for a U-curved relationship. A
M
possible explanation for the U-curve in SOL is the variation in study design over time. The older studies
D
(performed in the 1970s) more often used homogeneous samples (young, without comorbidity, without
15 sleep medications). They were also smaller (less patients) and less methodologically rigorous and
TE
therefore more susceptible to different types of bias. These methodological shortcomings might have
EP
led to higher effects. This is also in line with some of the other results of the multivariate analysis: in
studies of poorer methodological quality the interventions seemed to perform better. Studies that did
C
not report age or did not report the way they concealed treatment allocation showed better results than
AC
20 those who did. However, it might still be possible that the treatments offered in the 1970s were truly
more effective than those offered later on. The majority of these studies focused on either relaxation or
paradoxical intention. Maybe these treatments are indeed effective but have been pushed aside by the
current fashion of CBTI? After all, this also occurred in depression treatment where it has become state-
of-the-art to offer CBTI even though there is no conclusive proof that CBTI actually works better than
18
ACCEPTED MANUSCRIPT
other treatments [134]. More research is needed in which older insomnia treatments are directly
The treatment effects were not significantly related to the presence of comorbidity, the
allowance to use sleep medication, or to age. This is in line with previous reviews of insomnia
PT
5 treatments which argued that CBTI is effective for younger and older adults, for patients with and
RI
without comorbidities, for medication-free patients as well as for chronic hypnotic users [32].
An important limitation of this meta-analysis is that we only could include studies of which the
SC
results were published. There were indeed indications of publication bias. This suggests that the actual
effects of CBTI on insomnia might be smaller although still significant. Another limitation is that not all of
U
10 the studies used the same outcome measure. Differences in effects between outcomes might therefore
AN
reflect differences between studies rather than real differences in effect. Finally, many quality criteria
were not met by the studies which might mean that we overestimate the true effectiveness.
M
Despite the limitations we conclude that CBTI, either its components or the full package, is
D
effective in the treatment of insomnia. Face-to-face treatments and treatments of at least 4 sessions
15 seem to be more effective than self-help interventions or face-to-face interventions with fewer sessions.
TE
Otherwise the results seem to be quite robust (similar for patients with or without comorbid disease,
EP
younger or older patients, using or not using sleep medication). We need more high quality research,
comparing different treatment components directly with one another, to understand whether or not full
C
CBTI is more effective than its separate components and whether or not there are other (older)
AC
19
ACCEPTED MANUSCRIPT
Practice points
- Cognitive behavioral treatment for insomnia is effective with large overall effects on insomnia
severity, sleep efficiency, wake after sleep onset, and sleep onset latency. The magnitude of
these effects is in line with psychological treatments for other disorders.
- The smallest effects of cognitive behavioral treatment for insomnia are found for total sleep
time.
- Face-to-face treatments and treatments of at least 4 sessions seem to be more effective than
PT
self-help interventions or face-to-face interventions with fewer sessions.
RI
Research agenda
Future studies on cognitive behavioral treatment for insomnia should focus on:
- Determining which treatment components are essential for delivering cognitive behavioral
SC
treatment for insomnia
- Comparing face-to-face treatments for insomnia with their online counterpart
- Evaluating the optimal treatment length for cognitive behavioral treatment for insomnia
U
- Directly comparing earlier insomnia interventions (e.g. relaxation) to newer ones (e.g. CBTI)
AN
M
D
TE
C EP
AC
20
ACCEPTED MANUSCRIPT
References
1. Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep
Med Rev 2002; 6: 97–111.
2. Morin CM, LeBlanc M, Daley M, Grégoire JP, Merette C. Epidemiology of insomnia: prevalence,
self-help treatments, consultations, and determinants of help-seeking behaviors. Sleep Med 2006;
PT
7: 123–30.
3. Roth T, Ancoli-Israel S. Daytime consequences and correlates of insomnia in the United States:
RI
results of the 1991 National Sleep Foundation Survey. II. Sleep 1999; 22 (Suppl. 2):S 354–8.
4. Kyle SD, Espie CA, Morgan K. Quality of life and daytime functioning in insomnia. Behav Sleep Med
SC
2010; 8: 123–40.
5. Morin CM, Bélanger L, LeBlanc M, Ivers H, Savard J, Espie CA, et al. The natural history of
U
insomnia. a population-based 3-year longitudinal study. Arch Intern Med 2009; 169: 447–53.
6. Daley M, Morin CM, Le Blanc M, Grégoir JP, Savard J. The economic burden of insomnia: direct
AN
and indirect costs for individuals with insomnia syndrome, insomnia symptoms, and good
sleepers. Sleep 2009; 32 :55–64.
M
7. Daley M, Morin CM, LeBlanc M, Grégoir JP, Savard J, Baillargeon L. Insomnia and its relationship to
health-care utilization, work absenteeism, productivity and accidents. Sleep Med 2009; 10: 427–
D
38.
8. Budhiraja R, Roth T, Hudgel DW, Budhiraja P, Drake CL. Prevalence and polysomnographic
TE
correlates of insomnia comorbid with medical disorders. Sleep 2011; 34(7): 859-67.
9. Gallicchio L, Kalesan B. Sleep duration and mortality: a systematic review and meta-analysis. J
EP
11. Guo X, Zheng L, Wanga J, Zhang X, Zhang X, Li J, et al. Epidemiological evidence for the link
AC
between sleep duration and high blood pressure: A systematic review and meta-analysis. Sleep
Med 2013; 14: 324-32.
12. Laugsand LE, Vatten LJ, Platou C, Janszky I. Insomnia and the Risk of Acute Myocardial Infarction :
A Population Study. Circulation 2011; 124: 2073-81.
13. Redline S, Foody JA.Sleep Disturbances Time to Join the Top 10 Potentially Modifiable
Cardiovascular Risk Factors? Circulation 2011; 124: 2049-51.
21
ACCEPTED MANUSCRIPT
14. Suh S, Kim H, Yang H-C, Cho ER, Ku Lee S, Shin C. Longitudinal Course of Depression Scores with
and without Insomnia in Non-Depressed Individuals: A 6-Year Follow-Up Longitudinal Study in a
Korean Cohort. SLEEP 2013; 36(3): 369-76.
15. Okajima I, Komada Y, Nomura T, Nakashima K, Inoue Y. Insomnia as a risk for depression. A
longitudinal epidemiologic study on a Japanese rural cohort. J Clin Psychiatry 2012; 73(3): 377-83.
PT
16. Baglioni C, Battagliese G, Feige B, Spiegelhalder K, Nissen C, Voderholzer U, et al. Insomnia as a
predictor of depression: a meta-analytic evaluation of longitudinal epidemiological studies. J
RI
Affect Disord 2011; 135(1-3): 10-9.
17. Azevedo Da Silva M, Singh-Manoux A, Shipley MJ, Vahtera J, Brunner EJ, Ferrie JE, et al. Sleep
SC
duration and sleep disturbances partly explain the association between depressive symptoms and
cardiovascular mortality: the Whitehall II cohort study. J Sleep Res 2014; 23: 94–7.
18. Buscemi N, Vandermeer B, Friesen C, Bialy L, Tubman M, Ospina M, et al. The efficacy and safety
U
of drug treatments for chronic insomnia in adults: a meta-analysis of RCTs. J Gen Intern Med 2007;
19.
22: 1335–50. AN
Krystal AD. A compendium of placebo-controlled trials of the risks/benefits of pharmacological
treatments for insomnia: the empirical basis for U.S. clinical practice. Sleep Med Rev 2009; 13:
M
265–74.
20. Riemann D, Perlis ML.The treatments of chronic insomnia: a review of benzodiazepine receptor
D
agonists and psychological and behavioral therapies. Sleep Med Rev 2009; 13(3): 205–14.
TE
21. Edinger JD, Means MK. Cognitive-behavioral therapy for primary insomnia. Clin Psychol Rev 2005;
25: 539–58.
22. Edinger JD, Wolgemuth WK. The significance and management of persistent primary insomnia:
EP
the past, present and future of behavioral insomnia therapies. Sleep Med Rev 1999; 3: 101–18
23. Morin CM, Benca R. Chronic insomnia. Lancet 2012; 379(9821): 1129-41.
C
24. Morin CM, Espie CA. Insomnia: A Clinical Guide to Assessment and Treatment. New York, Kluwer
AC
22
ACCEPTED MANUSCRIPT
28. Morin CM, Culbert JP, Schwartz SM. Nonpharmacological interventions for insomnia: a meta-
analysis of treatment efficacy. Am J Psychiatry 1994; 151(8): 1172-80.
29. Murtagh R, Greenwood KM. Identifying effective psychological treatments for insomnia: a meta-
analysis. J Consult Clin Psychol 1995; 63(1): 79-9.
30. Harvey AG, Tang NKY. Cognitive behavior therapy for primary insomnia: can we rest yet? Sleep
PT
Med Rev 2003; 7(3): 237-62.
31. Montgomery P, Dennis J. A systematic review of non-pharmacological therapies for sleep
RI
problems in later life. Sleep Med Rev 2004; 8: 47–62.
32. * Morin CM, Bootzin RR, Buysse DJ, Edinger JD, Espie CA, Lichstein KL. Psychological and
SC
behavioral treatment of insomnia: update of the recent evidence (1998–2004). Sleep 2006; 29:
1398–414.
33. Siebern AT, Manber R. New developments in cognitive behavioral therapy as the first-line
U
treatment of insomnia. Psychol Res Behav Manag 2011; 4: 21–8.
34. AN
* Trauer JM, Qian MY, Doyle JS, Rajaratnam SM, Cunnington D. Cognitive behavioral therapy for
chronic insomnia: A systematic review and meta-analysis. Ann Intern Med 2015; 163(3): 191-204.
35. * Wu JQ, Appleman, ER, Salazar, RD, Ong JC. Cognitive behavioral therapy for insomnia comorbid
M
with psychiatric and medical Conditions. A meta-analysis. JAMA Intern Med 2015; 175(9): 1461-
72.
D
36. * Geiger-Brown JM, Rogers VE, Liu W, Ludeman EM, Downton KD, Diaz-Abad M. Cognitive
TE
behavioral therapy in persons with comorbid insomnia: A meta-analysis. Sleep Med Rev 2015; 23,
54-67.
37. * Smith MT, Perlis ML, Park A, Smith MS, Pennington J, Giles DE, et al. Comparative Meta-Analysis
EP
of pharmacotherapy and Behavior Therapy for Persistent Insomnia. Am J Psychiatry 2002; 159:5–
11.
C
38. * Mitchell MD, Gehrman P, Perlis M, Umscheid CA. Comparative effectiveness of cognitive
AC
behavioral treatment for insomnia: A systematic review. BMC Fam Pract 2012; 13: 40.
39. * Qaseem A, Kansagara D, Forciea MA, Cooke M, Denberg TD. Management of chronic insomnia
disorder in adults: A clinical practice guideline from the American College of Physicians. Ann
Intern Med in press.
40. Higgins JPT, Altman DG (eds). Assessing risk of bias in included studies, in Cochrane Handbook for
Systematic Reviews of Interventions. Edited by Higgins JPT, Green S. Chichester, UK, Wiley-
Blackwell, 2008, chap 8
23
ACCEPTED MANUSCRIPT
41. Cooper H, Hedges LV. The Handbook of Research Synthesis. New York, Russell Sage Foundation
Press, 1994
42. Hedges LV, Olkin I. Statistical Methods for Meta-Analysis. San Diego, CA, Academic Press, 1985
43. Lipsey MW, Wilson DB. The efficacy of psychological, educational and behavioural treatment.
Confirmation from meta-analysis. American Psychologist 1993; 48: 1181–209.
PT
44. Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based method of testing and adjusting for
publication bias in meta-analysis. Biometrics 2000; 56: 455–63.
RI
45. Alperson J, Biglan A. Self-administered treatment of sleep onset insomnia and the importance of
age. Behav Ther 1979; 10: 347-56.
SC
46. Arnedt JT, Conroy DA, Armitage R, Brower KJ. Cognitive-behavioural therapy for insomnia in
alcohol dependent patiens: a randomized controlled trial. Behav Res Ther 2011; 49: 227-33.
47. Arnedt JT, Cuddihy L, Swanson LM, Pickett S, Aikens J, Chervin RD. Randomized controlled trial of
U
telephone-delivered cognitive behavioral therapy for chronic insomnia. Sleep 2013; 36(3):353-62.
48. AN
Ascher LM, Turner RM. Paradoxical intention and insomnia: an experimental investigation. Behav
Res Ther 1979; 17:408-11.
49. Ascher LM, Turner RM. A comparison of two methods for the administration of paradoxical
M
intention. Behav Res Ther 1980; 18: 121-26.
50. Borkovec TD, Weerts TC. Effects of progressive relaxation on sleep disturbance: an
D
51. Bothelius K, Kyhle K, Espie C.A, Broman J-E. Manual-guided cognitive-behavioural therapy for
insomnia delivered by ordinary primary care personnel in general medical practice: a randomized
controlled effectiveness trial. J Sleep Res. 2013; 22(6): 688-96.
EP
52. Broomfield NM, Espie. An experimental investigation of putative mechanisms using subjective and
actigraphic measurement of sleep. Behav Cogn Psychother 2003; 31: 313-24.
C
53. Buysse DJ, Germain A, Moul DE, Franzen PL, Brar LK, Fletcher ME, et al. Efficacy of brief behavioral
AC
treatment for chronic insomnia in older adults. Arch Intern Med 2011; 171(10): 887-95.
54. Carr-Kaffashan L, Woolfolk RL. Active and placebo effects in treatment of moderate and severe
insomnia. J Consult Clin Psychol 1979; 47(6): 1072-80.
55. Creti L, Libman E, Bailes S, Fichten CS. Effectiveness of cognitive-behavioral insomnia treatment in
a community sample of older individuals: more questions than conclusions. J Clin Psychol Med
Settings 2005; 12(2): 153-64.
24
ACCEPTED MANUSCRIPT
56. Currie SR, Wilson KG, Pontefract AJ, deLaplante L. Cognitive-behavioral treatment of insomnia
secondary to chronic pain. J Consult Clin Psychol 2000; 68(3): 407-16.
57. Currie SR, Clark S, Hodgins DC, El-Guebaly N. Randomized controlled trial of brief cognitive-
behavioural interventions for insomnia in recovering alcoholics. Addiction 2004; 99: 1121-32.
58. Edinger JD, Wohlgemuth WK, Radtke RA. Cognitive behavioral therapy for treatment of chronic
PT
primary insomnia: a randomized controlled trial. JAMA. 2001; 285(14): 1856-64.
59. Edinger JD, Wohlgemuth WK, Krystal AD, Rice JR. Behavioral insomnia therapy for fibromyalgia
RI
patients. Arch Intern Med 2005; 165: 2527-35.
60. Edinger JD, Sampson WS. A primary care “friendly” cognitive behavioral insomnia therapy. SLEEP
SC
2003; 2: 177-82.
61. Edinger JD, Wohlgemuth WK, Radtke RA, Coffman CJ, Carney CE. Dose-response effects of
cognitive-behavioral insomnia therapy: a randomized clinical trial. Sleep 2007; 30(2): 203-12.
U
62. Edinger JD, Olsen MK, Stechuchak KM, Means MK, Lineberger MD, Kirby A, Carney CA. Cognitive
AN
behavioral therapy for patients with primary insomnia or insomnia associated predominantly with
mixed psychiatric disorders: a randomized clinical trial. Sleep 2009; 32(4): 499-510.
63. Ellis JG, Cushing, T, Germain A. Treating acute insomnia: A randomized controlled trial of a "single-
M
shot" of cognitive behavioral therapy for insomnia. Sleep 2015; 38(6): 971-78.
64. Epstein DR, Dirksen SR. Randomized trial of a cognitive-behavioral intervention for insomnia in
D
65. * Epstein DR, Sidani S, Bootzin RR, Belyea MJ. Dismantling multicomponent behavioral treatment
for insomnia in older adults: a randomized controlled trial. Sleep 2012; 35(6): 797-805.
66. Espie CA, Lindsay WR, Nrooks N, Hood EM, Turvey T. A controlled comparative investigation of
EP
psychological treatments for chronic sleep-onset insomnia. Behav Res Ther 1989; 27(1): 79-88.
67. Espie CA, Inglis SJ, Tessier S, Harvey L. The clinical effectiveness of cognitive behaviour therapy for
C
chronic insomnia: implementation and evaluation of a sleep clinic in general medical practice.
AC
25
ACCEPTED MANUSCRIPT
69. Espie CA, Fleming L, Cassidy J, Samuel L, Taylor LM, White CA, et al. Randomized controlled clinical
effectiveness trial of cognitive behavior therapy compared with treatment as usual for persistent
insomnia in patients with cancer. J Clin Oncol 2008; 26: 4651-8.
70. Espie CA, Kyle SD, Williams C, Ong JC, Douglas NJ, Hames P, Brown JSL. A randomized, placebo-
controlled trial of online cognitive behavioral therapy for chronic insomnia disorder delivered via
PT
an automated media-rich web application. Sleep 2012; 35(6): 769-81.
71. Freedman R, Papsdorf JD. Biofeedback and progressive relaxation treatment of sleep-onset
RI
insomnia: a controlled, all-night investigation. Biofeedback Self Regul 1976; 1(3): 253-71.
72. Friedman L, Benson K, Noda A, Zarcone V, Wicks DA, O'Connel K, et al. An actigraphic comparison
SC
of sleep restriction and sleep hygiene treatments for insomnia in older adults. J Geriat Psychiatry
Neurol 2000; 13: 17-27.
73. Harris J, Lack L, Kemp K, Wright H, Bootzin R. A randomized controlled trial of intensive sleep
U
retraining (ISR): a brief conditioning treatment for chronic insomnia. Sleep 2012; 35(1): 49-60.
74. AN
Haynes SN, Woodward S, Moran R, Alexander D. Relaxation treatment of insomnia. Behav Ther
1974; 5: 555-8.
75. Ho FYY, Chung KF, Yeung WF, Ng THY, Cheng SKW. Weekly brief phone support in self-help
M
cognitive behavioral therapy for insomnia disorder: Relevance to adherence and efficacy. Behav
Res Ther 2014; 63: 147-56.
D
76. Irwin MR, Olmstead R, Carrillo C, Sadeghi N, Breen EC, Witarama T. Cognitive behavioral therapy
TE
vs. Tai Chi for late life insomnia and inflammatory risk: a randomized controlled comparative
efficacy trial. Sleep 2014; 37(9): 1543-52.
77. Jacobs GD, Pace-Schott EF, Stickgold R, Otto MW. Cognitive behavior therapy and
EP
behavioral therapy for insomnia co-morbid with hearing impairment: a randomized controlled
AC
26
ACCEPTED MANUSCRIPT
81. Kaldo V, Jernelöv S, Blom K, Ljótsson B, Brodin M, Jörgensen, M. Guided internet cognitive
behavioral therapy for insomnia compared to a control treatment–a randomized trial. 2015; 71:
90-100.
82. Kapella MC, Herdegen JJ, Perlis ML, Shaver JL, Larson JL, Law JA, et al. Cognitive behavioral
therapy for insomnia comorbid with COPD is feasible with preliminary evidence of positive sleep
PT
and fatigue effects. Intern J COPD 2011; 6: 625–35.
83. Lacks P, Bertelson AD, Gans L, Kunkel J. The effectiveness of three behavioral treatments for
RI
different degrees of sleep onset insomnia. Behav Ther 1983a; 14: 593-605.
84. Lacks P, Bertelson AD, Sugerman J, Kunkel J. The treatment of sleep-maintenance insomnia with
SC
stimulus-control techniques. Behav Res Ther 1983b; 21(3): 291-95.
85. Lancee J, Van Den Bout J, Van Straten A, Spoormaker VI. Internet-delivered or mailed self-help
treatment for insomnia? A randomized waiting-list controlled trial. Behav Res Ther 2012; 50: 22-9
U
86. Lancee J, Eisma MC, van Straten A, Kamphuis JH. Sleep-related safety behaviors and dysfunctional
AN
beliefs mediate the efficacy of online CBT for insomnia: A randomized controlled trial. Cogn Behav
Ther 2015; 44(5): 406-22.
87. * Lancee J., van Straten A, Morina M, Kaldo V, Kamphuis JH. Guided online or face-to-face
M
cognitive behavioral treatment for Insomnia: A randomized wait-list controlled trial. Sleep 2016;
39: 183-91.
D
88. Lichstein KL, Peterson BA, Riedel BW, Means MK, Epperson MT, Aguillard RN. Relaxation to assist
TE
90. Lichstein KL, Riedel BW, Wilson NM, Lester KW, Aguillard RN. Relaxation and sleep compression
for late-life insomnia: a placebo-controlled trial. J Consult Clin Psychl 2001; 69(2): 227-39.
C
91. Lick JR, Heffler D. Relaxation training and attention placebo in the treatment of severe insomnia. J
AC
27
ACCEPTED MANUSCRIPT
94. McCrae CS, McGovern R, Lukefahr R, Stripling AM. Research evaluating brief behavioral sleep
treatments for rural elderly (RESTORE): a preliminary examination of effectiveness. Am J Geriatr
Psychiatry 2007; 15(11): 979-82.
95. Means MK, Lichstein KL, Epperson MT, Johnson CT. Relaxation therapy for insomnia: nighttime
and day time effects. Behav Res Ther 2000; 38: 665-78.
PT
96. Mimeault V, Morin CM. Self-help treatment for insomnia: Bibliotherapy with and without
professional guidance. J Consult Clin Psychol 1999; 67(4): 511-19.
RI
97. Morawetz D. Behavioral self-help treatment for insomnia: a controlled evaluation. Behav Ther
1989; 20: 365-79.
SC
98. Morin CM, Azrin NH. Behavioral and cognitive treatments of geriatric insomnia. J Consul Clin Psych
1988; 56(5): 748-53.
99. Morin CM, Kowatch RA, Barry T, Walton E. Cognitive-behavior therapy for late-life insomnia. J
U
Consult Clin Psychol 1993; 61(1): 137-46.
AN
100. Morin CM, Colecchi C, Stone J, Sood R, Brink D. Behavioral and pharmacological therapies for late-
life insomnia. JAMA 1999; 281: 991-9.
101. Morin CM, LeBlanc M, Savard J. Self-help treatment for insomnia: a randomized controlled trial.
M
Sleep 2005; 28(10): 1319-27.
102. Nicassio P, Bootzin R. A comparison of progressive relaxation and autogenic training as treatments
D
104. Pigeon WR, Moynihan J, Matteson-Rusby S, Jungquist CR, Xia Y, Tu X, et al. Comparative
effectiveness of CBT interventions for co-morbid chronic pain & insomnia: a pilot study. Behav Res
C
105. Riedel BW, Lichstein KL, Dwyer WO. Sleep compression and sleep education for older insomniacs:
self-help versus therapist guidance. Psychol Aging 1995; 10(1): 54-63.
106. Riedel BW, Lichstein KL, Peterson BA, Epperson MT, Means MK, Aguillard RN. A comparison of the
efficacy of stimulus control for medicated and nonmedicated insomniacs. Behav Modif 1998;
22(1): 3-28.
107. Riley WT, Mihm P, Behar A, Morin CM. A computer device to deliver behavioral interventions for
insomnia. Behav Sleep Med 2010; 8: 2-15.
28
ACCEPTED MANUSCRIPT
108. Ritterband LM, Thorndike FP, Gonder-Frederick LA, Magee JC, Bailey ET, Saylor DK, et al. Efficacy
of an internet-based behavioral intervention for adults with insomnia. Arch Gen Psychiatry 2009;
66(7): 692-8.
109. Ritterband LM, Bailey ET, Thorndike FP, Lord HR, Farrell-Carnahan L, Baum LD. Initial evaluation of
an Internet intervention to improve the sleep of cancer survivors with insomnia. Psycho-oncol
PT
2012; 21(7): 695-705.
110. Rybarczyk B, Lopez M, Benson R, Alsten C, Stepanski E. Efficacy of two behavioral treatment
RI
programs for comorbid geratric insomnia. Psychol Aging 2002; 17(2): 288-98.
111. Rybarczyk B, Stepanski E, Fogg L, Lopez M, Barry P, Davis A. A placebo-controlled test of cognitive-
SC
behavioral therapy for comorbid insomnia in older adults. J Consult Clin Psychol 2005; 73(6): 1164-
74.
112. Savard J, Simard S, Ivers H, Morin CM. Randomized study on the efficacy of cognitive-behavioral
U
therapy for insomnia secondary to breast cancer, part I: sleep and psychological effects. J Clin
Oncol 2005; 23: 6083-96. AN
113. Savard J, Ivers H, Savard MH, Morin CM. Is a video-based cognitive behavioral therapy for
insomnia as efficacious as a professionally administered treatment in breast cancer? Results of a
M
randomized controlled trial. Sleep 2014; 37(8): 1305-14.
114. Sivertsen B, Omvik S, Pallesen S, Bjorvatn B, Havik OE, Kvale G, et al. Cognitive behavioral therapy
D
vs zopiclone for treatment of chronic primary insomnia in older adults. JAMA 2006; 295: 2851-8.
TE
115. Smith, MT, Finan PH, Buenaver LF, Robinson M, Haque U, Quain A, et al. Cognitive–behavioral
therapy for insomnia in knee osteoarthritis: A randomized, double-blind, active placebo–
controlled clinical trial. Arthritis Rheum 2015; 67(5): 1221-33.
EP
116. Soeffing JP, Lichstein KL, Nau SD, McCrae CS, Wilson NM, Aguillard RN, et al. Psychological
treatment of insomnia in hypnotic-dependant older adults. Sleep Med 2008; 9:165-171.
C
117. Stanton HE. Hypnotic relaxation and the reduction of sleep onset insomnia. Int J Psychosom 1989;
AC
36(1-4): 64-8.
118. Steinmark SW, Borkovec TD. Active and placebo treatment effects on moderate insomnia under
counterdemand and positive demand instructions. J Abnorm Psychol 1974; 83(2): 157-63.
119. Ström L, Petterson R, Andersson G. Internet-based treatment for insomnia: a controlled
evaluation. J Consult Clin Psychol 2004; 72(1): 113-20.
29
ACCEPTED MANUSCRIPT
120. Swift N, Stewart R, Andiappan M, Smith A, Espie CA, Brown JSL. The effectiveness of community
day-long CBT-I workshops for participants with insomnia symptoms: a randomised controlled trial.
J Sleep Res 2011; 21(3): 270-80.
121. Talbot LS, Maguen S, Metzler TJ, Schmitz M, McCaslin SE, Richards A, et al. Cognitive behavioral
therapy for insomnia in posttraumatic stress disorder: a randomized controlled trial. Sleep
PT
2014; 37(2): 327-41.
122. Taylor DJ, Schmidt-Nowara W, Jessop CA, Ahearn J. Sleep restriction therapy and hypnotic
RI
withdrawal versus sleep hygiene education in hypnotic using patients with insomnia. J Clin Sleep
Med 2010; 6(2): 169-75.
SC
123. Taylor DJ, Zimmerman MR, Gardner CE, Williams JM, Grieser EA, Tatum JI, et al. A pilot
randomized controlled trial of the effects of cognitive-behavioral therapy for insomnia on sleep
and daytime functioning in college students. Behav Ther 2014; 45(3): 376-89.
U
124. Turner RM, Ascher LM. Controlled comparison of progressive relaxation, stimulus control, and
AN
paradoxical intention therapies for insomnia. J Consult Clin Psychol 1979; 47(3): 500-8.
125. van Straten A, Cuijpers P, Smit F, Spermon M, Verbeek I. Self-help treatment for insomnia through
television and book: a randomized trial. Patient Educ Couns 2009; 74: 29-34.
M
126. van Straten A, Emmelkamp J, de Wit J, Lancee J, Andersson G, van Someren EJ, et al. Guided
Internet-delivered cognitive behavioural treatment for insomnia: a randomized trial. Psychol Med
D
2013: epub.
TE
127. Vincent N, Lewycky S. Logging on for better sleep: RCT of the effectiveness of online treatment for
insomnia. SLEEP 2009; 32(6): 807-15.
128. Waters WF, Hurry MJ, Binks PG, Carney CE, Lajos LE, Fuller KH, et al. Behavioral and hypnotic
EP
treatments for insomnia subtypes. Behav Sleep Med 2003; 1(2): 81-101.
129. Woolfolk RL, McNulty TF. Relaxation treatment for insomnia: a component analysis. J Consult Clin
C
130. Wu R, Bao J, Zhang C, Deng J, Long C. Comparison of sleep condition and sleep-related
psychological activity after cognitive-behavior and pharmalogical therapy for chronic insomnia.
Psychother Psychosom 2006; 75: 220-8.
131. Zwart CA, Lisman SA. Analysis of stimulus control treatment of sleep-onset insomnia. J Consult Clin
Psychol 1979; 47(1): 113-8.
132. Morin CM, Belleville G, Belanger L, Ivers H. The insomnia severity index: psychometric indicators
to detect insomnia cases and evaluate treatment response. Sleep 2011; 34(5): 601-8.
30
ACCEPTED MANUSCRIPT
133. Buysse DJ, Reynolds III CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a
new instrument for psychiatric research and practice. Psychiatry Res 1989; 28: 193-213.
134. Cuijpers P, van Straten A, Andersson G, Van Oppen P. Psychotherapy for depression in adults: a
meta-analysis of comparative outcome studies. J Consult Clin Psychol 2008; 76(6): 909-22.
135. Acarturk C, Cuijpers P, van Straten A, de Graaf R. Psychological treatment of social anxiety
PT
disorder: a meta-analysis. Psychol Med 2009; 39(2): 241-54.
136. Sánchez-Meca J, Rosa-Alcázar AI, Marín-Martínez F, Gómez-Conesa A. Psychological treatment of
RI
panic disorder with or without agoraphobia: A meta-analysis. Clin Psychol Rev 2010; 30: 37–50.
137. Buysse DJ, Ancoli-Israel S, Edinger JD, Lichstein KL, Morin CM. Recommendations for a standard
SC
research assessment of insomnia. Sleep. 2006; 29(9): 1155-73.
138. Harvey AG, Bélanger L, Talbot L, Eidelman P, Beaulieu-Bonneau S, Fortier-Brochu E, et al.
Comparative efficacy of behavior therapy, cognitive therapy, and cognitive behavior therapy for
U
chronic insomnia: a randomized controlled trial. J Consult Clin Psychol 2014; 82(4): 670-683.
AN
139. Kyle SD, Aquino MR, Miller CB, Henry AL, Crawford MR, Espie CA, et al. Towards standardisation
and improved understanding of sleep restriction therapy for insomnia disorder: a systematic
examination of CBT-I trial content. Sleep Med Rev 2015; 23: 83-88.
M
140. Cuijpers P, Donker T, van Straten A, Li J, Andersson G. Is guided self-help as effective as face-to-
face psychotherapy for depression and anxiety disorders? A systematic review and meta-analysis
D
141. Andersson G, Cuijpers P, Carlbring P, Riper H, Hedman E. Guided internet-based vs. face-to-face
cognitive behavior therapy for psychiatric and somatic disorders: a systematic review and meta-
analysis. World Psychiatry 2014; 13: 288-295.
EP
142. Blom K, Tarkian Tillgren H, Wiklund T, Danlycke E, Forssén M, Söderström A, et al. Internet-vs.
group-delivered cognitive behavior therapy for insomnia: a randomized controlled non-inferiority
C
31
ACCEPTED MANUSCRIPT
Table 1. Characteristics of the included studies on (elements of) CBT for insomnia
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
PT
45
Alperson 1979 Comm SOL ≥ 30m Allowed 17-54 Relax+SC Group 2 7 WL 7
RI
46
Arnedt 2011 Care SOL ≥ 30m, ≥ 3 n/w, ≥ 1mo + ISI ≥ 8 Alcohol 18-65 CBTI Indiv 4 9 Plac 8
47
SC
Arnedt 2013 Care wake time > 60m+ SE < 85% Excluded 18-65 CBTI Phone 4-8 18 Info 15
48
Ascher 1979 Comm SOL ≥ 60m, ≥ 3 n/w Excluded 18+ Pdox Indiv 4 8 No tx 9
U
49
Ascher 1980 Comm No criteria Excluded 18+ Pdox A Indiv 4 10 WL 10
AN
Pdox B Indiv 4 10
50
Borkovec 1976 Univ Average SOL ≥ 30 mins Allowed - Relax Group 4 12 No txt 12
M
51
Bothelius 2013 Care Poor sleep > 1m month + daytime imp Excluded 18+ CBTI Group 5 32 WL 34
D
52
Broomfield 2003 Univ SOL ≥ 30m, > 3 n/w + PSQI > 5 Excluded 16-65 Pdox SH n/a 17 No tx 17
Buysse 2011
53
Care
TE
DSM-IV and ICSD-2 insomnia diagnosis Allowed elderly Behav Indiv 2 42 Info 40
54
EP
Carr-Kaffashan 1979 Comm Average SOL ≥ 30 m, ≥ 6 mo Allowed 18+ Relax Indiv 4 16 Plac 14
55
Creti 2005 Comm No criteria Excluded 55+ Relax SH n/a 14 No tx 13
C
56
Currie 2000 Care DSM-III insomnia diagnosis with SIS-D and ICSD diagnosis Pain < 60 CBTI Group 7 32 WL 28
AC
57
Currie 2004 Care SOL ≥ 30 m, ≥ 3 n/w Alcohol 18+ CBTI Indiv 5 20 WL 20
CBTI SH n/a 20
58
Edinger 2001 Comm mean WASO ≥ 60 m, ≥ 6 mo Excluded 40-80 behavioral Indiv 6 25 Plac 25
Relax Indiv 6 25
1
ACCEPTED MANUSCRIPT
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
59
Edinger 2005 Comm mean WASO ≥ 60 m Fibromyalgia 21-65 Behavioral Indiv 6 18 No tx 11
PT
60
Edinger 2003 Primary Difficulty initiating or maintaining sleep ≥ 1 mo + daytime imp Excluded - BT Indiv 2 10 plac 10
RI
care
SC
behavioral Indiv 2 18
behavioral Indiv 4 24
U
behavioral Indiv 8 17
AN
62
Edinger 2009 Care mean SOL + WASO ≥ 60 m Excl (PI) - behavioral Indiv 4 20 Info 40
M
63
Ellis, 2015 Comm Difficulty initating or maintaining sleep or early morning awakening, excluded - CBTI Indiv 1 20 WL 20
≥ 3 n/w, < 3 mo
D
64
Epstein 2007 Comm SOL or WASO ≥ 30 m, ≥ 3 n/w, ≥ 2 weeks + daytime imp
TE Cancer 18+ behavioral Group 4 40 Info 41
65
Epstein 2012 Comm SOL or WASO ≥ 45 m, ≥ 3 n/w, ≥ 6 mo+ day-time imp Excluded 55+ PE+SR Group 4 44 WL 50
EP
PE+SC Group 4 44
PE+SC+SR Group 4 41
C
66
Espie 1989 Care Mean SOL ≥ 30 m, ≥ 1 year Excluded - SC Group 8 14 No tx 13
AC
Relax Group 8 14
Pdox Group 8 15
67
Espie 2001 Care ICSD difficulty falling/maintaining sleep, ≥ 4 n/w, ≥ 3 mo + PSQI ≥ 5 Excluded - CBTI Group 6 74 WL 65
68
Espie 2007 Care ICSD / DSM-IV criteria of insomnia Allowed - CBTI Group 5 107 No tx 94
2
ACCEPTED MANUSCRIPT
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
69
Espie 2008 Care SOL or WASO ≥ 30 m, ≥ 3 n/w, ≥ 3 mo + PSQI ≥ 5 Cancer 18+ CBTI Group 5 100 No tx 50
PT
70
Espie 2012 Comm poor sleep ≥ 3 n/w, ≥ 3 mo + daytime imp + SE < 80% Excluded 18+ CBTI SH n/a 55 WL 54
RI
71
Freedman 1976 Univ SOL ≥ 60 m, ≥ 4 n/w, ≥ 6 mo Excluded - Relax Indiv 6 6 Info 6
72
SC
Friedman 2000 Comm SE < 80%, SOL > 30m, TST < 6hr, WASO > 30 m, ≥ 5 n/2w Excluded 55+ SC+SR Indiv 5 16 Info 11
73
Harris 2012 Comm SOL > 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 18-65 SC Indiv 5 20 Info 20
U
74
Haynes 1974 Univ NR - 18-21 Relax Group 6 7 plac 7
AN
75
Ho 2014 Comm SOL or WASO ≥ 3 n/w, ≥ 3 mo + daytime imp Allowed 18+ CBTI SH+ n/a 103 WL 105
M
76
Irwin 2014 Comm SOL or WASO ≥ 3 n/w, ≥ 3 mo + daytime imp Excluded 55+ PE+SC+CT group 16 50 info 25
D
+relax
Jacobs 2004
77
Comm
TE
SOL > 60m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 25-64 CBTI Indiv 4 15 Plac 15
78
Jansson 2012 Care SOL or WASO > 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Hearing 18-65 CBTI Indiv 7 17 WL 15
EP
79
Jernelov 2012 Comm ISI > 10 + poor sleep ≥ 4 weeks Excluded 18+ CBTI SH N/A 44 WL 44
CBTI SH N/A 45
C
AC
Jungquist 2010 80 Care SOL or WASO > 30m, > 3 n/w, > 6 mo Pain 25+ CBTI Indiv 8 19 Plac 9
81
Kaldo 2015 Comm Difficulty initiating or maintaining sleep + daytime imp + ISI > 10 Excluded 18+ CBTI SH n/a 73 Plac 75
82
Kapella 2011 Comm Difficulty initiating or maintaining sleep, waking up too early or poor COPD 45+ CBTI Indiv 6 9 plac 9
3
ACCEPTED MANUSCRIPT
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
quality sleep
PT
83
Lacks 1983a Comm SOL ≥ 30m, ≥ 1 n/w, ≥ 6 mo Excluded 17-59 Relax Group 4 19 Plac 16
SC Group 4 15
RI
Pdox Group 4 14
SC
84
Lacks 1983b Comm WASO ≥ 30m, ≥ 1 n/w, ≥ 6 mo Excluded 17-59 SC Group 4 7 Plac 8
85
Lancee 2012 Comm SE < 85% + SLEEP-50 ≥ 19 Allowed 18+ CBTI SH N/A 205 WL 202
U
CBTI SH N/A 216
AN
86
Lancee 2015 Comm SOL or WASO > 30m, ≥ 3 n/w, ≥ 3 mo + daytime imp + ISI > 10 Excluded 18+ CBTI SH n/a 36 WL 27
87
M
Lancee 2016 Comm SOL or WASO ≥ 30m, ≥ 3 n/w, ≥ 3 mo + ISI ≥ 10 Excluded 18+ CBTI SH n/a 30 WL 30
Indiv 6 30
D
88
Lichtstein 1999 Comm SOL or WASO ≥ 30m, ≥ 3 n/w, ≥ 6 mo Excluded - Relax Indiv 3 10 WL 10
Lichtstein 2000
89
Comm
TE
SOL or WASO ≥ 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Illness 58+ Relax+SC Indiv 4 23 WL 21
90
Lichtstein 2001 Comm SOL or WASO ≥ 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 59+ Relax Indiv 6 27 Plac 23
EP
SR Indiv 6 24
C
91
Lick 1977 Comm Average SOL ≥ 50m Excluded - Relax+ Indiv 6 10 No tx 10
Relax- Indiv 6 10
AC
92
Lovato 2014 Comm WASO > 30m, ≥ 3 n/w, > 6 mo + daytime imp Exluded Older CBTI Group 4 86 WL 32
93
Manber 2008 Comm SOL or WASO ≥ 30m, TST ≤ 6.5hrs, ≥ 3 n/w MDD 18-75 CBTI Indiv 7 15 Plac 15
94
McCrae 2007 Comm SOL > 31m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 65+ SC+SR Indiv 2 11 info 11
4
ACCEPTED MANUSCRIPT
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
95
Means 2000 Univ SOL or WASO ≥ 30m + spoor sleep ≥ 2 mo + daytime imp Excluded - Relax Indiv 3 28 No tx 29
PT
96
Mimeault 1999 Comm SOL or WASO ≥ 30m, ≥ 3 n/w, ≥ 1 mo + daytime imp Excluded 18+ CBTI SH N/A 18 WL 18
RI
CBTI SH N/A 18
97
Morawetz 1989 Comm SOL or WASO > 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 18-60 SR+relax SH N/A 16 WL 18
SC
SR+relax Group 5 16
98
Morin 1988 Comm WASO > 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 55+ SC Group 6 9 WL 10
U
99
Morin 1993 Comm WASO > 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 60+ CBTI Group 8 12 WL 12
AN
100
Morin 1999 Comm SOL or WASO ≥ 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 55+ CBTI Group 8 18 Plac 20
M
101
Morin 2005 Comm Poor sleep ≥ 3 n/w Allowed 18+ CBTI SH n/a 96 No tx 96
D
102
Nicassio 1974 Comm Average SOL ≥ 30m Allowed - RelaxA Indiv 4 8 No tx 7
TE RelaxB Indiv 4 8
103
Norell-Clarke 2015 Comm Insomnia symptoms > 3 n/w + ISI > 10 Depression - CBTI Group 4 32 Plac 32
EP
104
Pigeon 2012 Care SOL or WASO > 30 m, ≥ 3 n/w, ≥ 6 mo Pain 35-75 CBTI Indiv 10 6 WL 4
105
Riedel 1995 Comm SOL > 30m, ≥ 3 n/w, ≥ 12 mo Excluded 60+ PE+SR Group 4 25 WL 25
C
106
AC
Riedel 1998 Comm SOL or WASO or early awakening > 30m, ≥ 3 n/w, ≥ 6 mo Excluded 19-80 SC Indiv 2 11 WL 10
107
Riley 2010 Comm SOL or WASO or early awakening > 30m, ≥ 3 n/w, ≥ 6 mo Excluded 18-65 SC+SR SH n/a 57 Info 33
108
Ritterband 2009 Comm Poor sleep ≥ 3 n/w, ≥ 6 mo + daytime imp Allowed 18-65 CBTI SH n/a 22 WL 23
109
Ritterband 2012 Comm Poor sleep ≥ 3 n/w, ≥ 6 mo + daytime imp + average TST ≤ 6.5 hr Cancer 21+ CBTI SH n/a 14 WL 14
5
ACCEPTED MANUSCRIPT
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
110
Rybarczyk 2002 Care SOL ≥ 45m or WASO ≥ 60m or TST ≤ 5 hr, ≥ 3 n/w Illnesses 55+ CBTI Group 8 16 WL 13
PT
Relax SH n/a 18
111
Care SOL ≥ 30m/ WASO ≥ 60m/ TST ≤ 6.5 hr, ≥ 3 n/w, ≥ 6 mo + daytime Ilness 55+ CBTI Group 8 46 Plac 46
RI
Rybarczyk 2005
imp
SC
112
Savard 2005 Comm SOL or WASO > 30m, SE < 85%, ≥ 3 n/w, ≥ 6 mo + daytime imp Cancer 18+ CBTI Group 8 27 WL 30
113
Savard 2014 Care ISI ≥ 8 or ≥ 2 nights of sleepmedication in last 2 weeks Cancer 18-75 CBTI Indiv 6 81 No tx 81
U
CBTI SH n/a 80
AN
114
Sivertsen 2006 Comm Poor sleep ≥ 3 mo + daytime imp Allowed 55+ CBTI Indiv 6 18 Plac 12
115
M
Smith 2015 Care SOL or WASO > 30m, ≥ 2 n/w, > 1 mo Knee - CBTI Indiv 8 50 Plac 50
osteoarthritis
D
116
Soeffing 2008 Comm SOL or WASO > 30m, ≥ 3 n/w, ≥ 6 mo + daytime imp Excluded 50+ CBTI Indiv 8 20 Plac 27
Stanton 1989
117
Comm Average SOL > 30m, ≥ 6 mo
TE Allowed - Relax Indiv 4 15 Plac 15
SC Indiv 4 15
EP
118
Steinmark 1974 Univ Average SOL > 30m, ≥ 6 mo Allowed - RelaxA Group 4 12 WL 12
RelaxB Group 4 12
C
119
Strom 2004 Comm SOL / WASO/ early awakening > 30m, ≥ 3 n/w, ≥ 3 mo + daytime Excluded 18+ CBTI SH n/a 54 WL 55
AC
imp
120
Swift 2012 Comm no in- or exclusion criteria Allowed 18+ CBTI Group 1day 75 WL 76
121
Talbot 2014 Comm ISI > 14 + difficulty initiating or maintaining sleep + daytime imp PTSD 18-65 CBTI Indiv 8 29 WL 16
treatment
6
ACCEPTED MANUSCRIPT
study ment Definition insomnia morbidity Age interv format sess EXP group CTR
122
Taylor 2010 Care Average SE < 85% , ≥ 6 mo + regular use of sleep medication Excluded 18+ SR Indiv 8 24 WL 22
PT
123
Taylor 2014 Univ SOL or WASO > 30m, ≥ 3 n/w, ≥ 3 mo + daytime imp Allowed 18-27 CBTI Indiv 6 17 WL 17
RI
124
Turner 1979 Comm Not reported Allowed - Relax Indiv 4 10 No tx 10
Pdox Indiv 4 10
SC
PR Indiv 4 10
125
Van Straten 2009 Comm SOL / WASO / early awakening > 30m, ≥ 3 n/w, ≥ 1 mo Allowed 18+ CBTI SH n/a 126 WL 121
U
126
AN
Van Straten 2013 Comm SOL > 30m, ≥ 3 n/w, ≥ 3 mo Allowed 18+ CBTI SH n/a 59 WL 59
127
Vincent 2009 Comm SOL or WASO > 30m, ≥ 4 n/w, ≥ 6 mo + daytime imp Allowed 18+ CBTI SH n/a 59 WL 59
M
128
Waters 2003 Comm SOL and (WASO > 30 mins or NWAK > 3) for ≥ 4 nights week for ≥ Excluded 18-59 Relax Indiv 4 13 Info 16
D
129
Woolfolk 1983 Comm Average SOL ≥ 50m, ≥ 6 mo TE Excluded 18+ RelaxA Group 4 8 WL 7
RelaxB Group 4 10
RelaxC Group 4 9
EP
130
Wu 2006 Comm SOL or WASO ≥ 30m, ≥ 6 mo + daytime imp Excluded - CBTI Group 16 19 Plac 19
C
131
Zwart 1979 Univ Average SOL ≥ 30m Allowed - SC Group 4 9 WL 8
AC
Abbreviations: AASM = American Academy of Sleep Medicine; AT = autogenic training; CD = cognitive distraction; Comm = Community; CMI =
comorbid insomnia; CT = cognitive therapy; DSISD = Duke Structured Interview for Sleep Disorders; F2F = face-to-face; ICSD = International
Classification of Sleep Disorders; ISI = insomnia severity index; NEXP = number of patients in experimental condition; Nctrl = number of patient in
control condition; NWAK = number of awakenings; NR = not reported; PE = psycho-education; PI = primary insomnia; PR = progressive
relaxation; PSQI = Pittsburgh Sleep Quality Index; RDC = Research Diagnostic Criteria; SC = stimulus control; SE = sleep efficiency; SH =
sleephygiene; SIS-D = Structured Interview for Sleep Disorders for DSM-III-R ; SR = sleep restriction; TST = total sleep time; WL = waitlist;
7
ACCEPTED MANUSCRIPT
PT
31 0.92 (0.79-1.06) 51 (19~67) 2.07
Without 7 outliers1
Sleep efficiency (SE) 79 0.71 (0.61-0.82) 70 (61~75) 2.60
RI
Without 14 outliers2 65 0.68 (0.60-0.74) 34 (5~51) 2.70
Pittsburgh sleep quality index3 19 0.65 (0.51-0.79) 39 (0~64) 2.82
SC
Wake after sleep onset (WASO) 71 0.63 (0.53-0.73) 60 (46~68) 2.91
Without 9 outliers4 62 0.66 (0.57-0.74) 40 (14~55) 2.78
U
Sleep onset latency (SOL) 108 0.57 (0.50-0.65) 48 (33~58) 3.18
Without 9 outliers5 99 0.55 (0.48-0.61) 27 (3~43) 3.31
AN
Sleep quality (SQ) 40 0.40 (0.24-0.56) 74 (64~80) 4.50
Without 6 outliers6 34 0.45 (0.31-0.59) 49 (17~65) 4.00
M
Number of awakenings (NWAK) 36 0.28 (0.16-0.40) 29 (0~52) 6.41
Without 2 outliers7 34 0.28 (0.17-0.38) 11 (0~42) 6.41
D
Total sleep time (TST) 91 0.16 (0.08-0.24) 47 (30~58) 11.11
1
Without 9 outliers8
2
82
3
TE
0.17 (0.11-0.24)
4
14 (0~35)
5
10.42
6 7
3 lower and 4 higher; 7 lower and 7 higher; no outliers present; 6 lower and 3 higher; 3 lower and 6 higher; 4 lower and 2 higher; 1 lower and 1
8
higher; 6 lower and 3 higher; * p < 0.05; ** p < 0.01; 95% CI = 95% confidence interval; Nc = number of comparisons; NNT = number needed to treat
C EP
AC
1
ACCEPTED MANUSCRIPT
Table 3. Post-test effects of insomnia treatment on SOL, SE and ISI: subgroup analyses
Sleep Onset Latency (SOL) Sleep Efficiency (SE) Insomnia Severity Index (ISI)
2 2 2
NC g (95% CI) I (95% CI p NC g (95% CI) I (95% CI p NC g (95% CI) I (95% CI p
TREATMENT VARIABLES
PT
Full CBTI
Yes 44 0.47 (0.41 – 0.53) 51 (27~65) 0.25 47 0.63 (0.56-0.69) 70 (58~77) 0.41 34 0.82 (0.74-0.91) 75 (65~81) 0.98
No 64 0.61 (0.52 - 0.70) 44 (20~58) 32 0.69 (0.58-0.79) 70 (55~78) 4 1.03 (0.79-1.26) 48 (0~81)
RI
Type of treatment
CBTI 44 0.47 (0.41 – 0.53) 51 (27~65) 0.03 47 0.63 (0.56-0.69) 70 (58~77) < 0.01 34 0.82 (0.74-0.91) 75 (65~81) 0.67
5)
Behavioral 12 0.38 (0.20 - 0.55) 45 (0~70) 13 0.68 (0.50-0.85) 67 (32~80) 2 0.90 (0.53—1.26) 80 n/a
SC
Relaxation 22 0.63 (0.46 – 0.80) 35 (0~60) 5 0.18 (-0.09-0.45) 0.0 (0~64) - - - -
5)
Other 30 0.72 (0.59 – 0.84) 42 (0~62) 14 0.87 (0.71 - 1.03) 72 (47~82) 2 1.12 (0.81-1.42) 0 n/a
Treatment format
U
F2F individual 50 0.61 (0.51 – 0.71) 50 (27~63) 0.07 37 0.66 (0.56-0.77) 47 (17~64) 0.03 14 1.11 (0.94-1.28) 68 (37~80) 0.21
F2F group 33 0.59 (0.49 – 0.68) 42 (3~61) 17 0.90 (0.78-1.01) 79 (65~85) 9 0.98 (0.82-1.13) 48 (0~74)
AN
Selfhelp 25 0.42 (0.34 – 0.49) 42 (0~63) 25 0.52 (0.44-0.59) 71 (54~80) 15 0.68 (0.58-0.79) 78 (62~85)
1
Number of sessions
1-4 41 0.49 (0.38-0.60) 44 (13~61) 0.06 16 0.54 (0.39-0.69) 12 (0~52) 0.01 4 1.03 (0.79-1.27) 40 (0~79) 0.44
5-6 26 0.66 (0.56-0.76) 49 (11~67) 22 0.84 (0.73-0.95) 76 (63~83) 10 1.12 (0.96-1.29) 52 (0~75)
M
7 or more 16 0.68 (0.52-0.84) 43 (0~67) 16 0.88 (0.72-1.04) 57 (13~74) 9 0.89 (0.68-1.10) 74 (40~85)
n/a (selfhelp) 25 0.42 (0.34 – 0.49) 42 (0~63) 25 0.52 (0.44-0.59) 71 (54~80) 15 0.68 (0.58-0.79) 78 (62~85)
D
PATIENT VARIABLES
Population
Community
Other
79
29
0.47 (0.41 – 0.53)
0.64 (0.54 – 0.74)
51 (34~62)
29 (0~54)
TE
0.10 59
20
0.60 (0.54-0.66)
0.77 (0.66-0.88)
70 (61~77)
64 (37~77)
0.10 26
12
0.82 (0.73-0.91)
0.93 (0.78-1.08)
78 (68~84) 0.59
54 (0~74)
Comorbidity
Yes 20 0.63 (0.52 – 0.75) 36 (0~62) 0.38 20 0.87 (0.75-0.99) 55 (16~72) 0.03 14 0.90 (0.76-1.05) 68 (37~80) 0.65
EP
No 88 0.49 (0.43 – 0.55) 49 (33~60) 59 0.58 (0.52-0.64) 70 (61~77) 24 0.82 (0.73-0.91) 77 (65~83)
Sleep medication
Allowed 51 0.46 (0.40 – 0.52) 49 (25~63) < 0.01 41 0.58 (0.52-0.65) 70 (57~77) 0.22 25 0.77 (0.68-0.86) 76 (64~83) 0.13
C
Not allowed 47 0.59 (0.49 – 0.68) 48 (22~62) 35 0.76 (0.66-0.87) 69 (55~77) 11 1.06 (0.89-1.22) 62 (10~79)
5)
Not reported 10 0.93 (0.69 - 1.17) 0 (0~53) 3 0.99 (0.65-1.32) 0 (0~73) 2 1.58 (0.91-2.25) 0 n/a
AC
3
Age
Younger adults only 21 0.70 (0.57 – 0.84) 0 (0~41) 0.02 13 0.99 (0.84-1.15) 0 (0~49) < 0.01 10 1.02 (0.84-1.19) 72 (38~84) 0.16
All adults 36 0.42 (0.35 – 0.49) 52 (24~66) 29 0.55 (0.49-0.63) 71 (56~79) 19 0.73 (0.63-0.84) 78 (65~85)
Older adults only 24 0.52 (0.40 – 0.64) 48 (7~67) 27 0.82 (0.71-0.94) 72 (58~80) 6 1.11 (0.92-1.31) 0 (0~61)
Not reported 27 0.63 (0.51 – 0.75) 52 (17~68) 10 0.34 (0.18-0.50) 0 (0~53) 3 0.68 (0.41-0.95) 35 (0~81)
1
ACCEPTED MANUSCRIPT
STUDY VARIABLES
Year of publication
16 1.03 (0.80 – 1.27) 0 (0~45) < 0.01 - - - - < 0.01 - - - - 0.82
1970-1979
16 0.73 (0.53 - 0.92) 52 (0~72) - - - - - - - -
1980-1989 5)
6 0.35 (0.07 – 0.64) 0 (0~61) 7 0.56 (0.30-0.83) 55 (0~79) 2 1.15 (0.66-1.64) 0 n/a
1990-1999
PT
37 0.41 (0.32 - 0.49) 9 (0~39) 38 0.46 (0.38-0.55) 58 (37~70) 8 0.71 (0.54-0.89) 82 (62~89)
2000-2009
33 0.53 (0.46 – 0.60) 64 (46~75) 34 0.77 (0.70-0.85) 74 (62~80) 28 0.87 (0.78-0.96) 73 (58~80)
2010+
Type of control
RI
No treatment 21 0.66 (0.54 – 0.77) 53 (13~70) 0.09 9 0.57 (0.44-0.70) 77 (50~86) 0.27 3 0.63 (0.45-0.81) 88 (50~94) 0.03
Waitlist 54 0.49 (0.43 – 0.56) 53 (33~65) 43 0.68 (0.61-0.75) 76 (68~82) 25 0.95 (0.85-1.04) 76 (64~83)
Placebo 20 0.47 (0.34 – 0.61) 50 (5~69) 15 0.60 (0.45-0.74) 51 (0~72) 8 0.79 (0.60-0.97) 32 (0~69)
SC
5)
Info/other 13 0.45 (0.28 – 0.62) 0 (0~49) 12 0.59 (0.42-0.77) 0.0 (0~50) 2 0.41 (-0.07-0.90) 0 n/a
Allocation sequence
adequate 43 0.49 (0.43-0.55) 62 (44~72) 0.51 46 0.67 (0.60-0.73) 76 (67~81) 0.13 27 0.86 (0.77-0.94) 78 (68~84) 0.53
U
not reported 65 0.57 (0.48-0.65) 33 (4~50) 33 0.56 (0.44-0.67) 53 (25~68) 11 0.79 (0.60-0.98) 56 (0~76)
Allocation concealed
AN
Yes 26 0.39 (0.32-0.47) 53 (2-~69) < 0.01 26 0.58 (0.50-0.65) 73 (58~81) 0.54 14 0.67 (0.56-0.78) 74 (52~84) 0.08
Not reported 82 0.63 (0.56-0.70) 38 (16~53) 53 0.71 (0.63-0.79) 67 (55~75) 24 1.03 (0.92-1.15) 67 (46~77)
Study drop-out
5)
Unknown 11 0.71 (0.46-0.97) 61 (7~78) 0.62 2 0.13 (-0.44-0.69) 0 n/a 0.21 - - - - 0.25
M
0-9% 34 0.58 (0.47-0.70) 14 (0~44) 28 0.66 (0.54-0.78) 10 (0~44) 13 0.90 (0.73-1.08) 65 (26~79)
10-19% 47 0.50 (0.43-0.57) 57 (37~68) 32 0.68 (0.61-0.76) 80 (73~85) 17 0.95 (0.94-1.05) 77 (62~85)
20% or more 16 0.46 (0.34-0.57) 52 (0~71) 17 0.55 (0.44-0.66) 73 (54~82) 8 0.61 (0.46-0.75) 68 (13~83)
D
ITT analyses
Yes 51 0.51 (0.44-0.57) 62 (47~71) 0.67 50 0.66 (0.60-0.73) 75 (67~80) 0.09 33 0.83 (0.75-0.91) 74 (63~81) 0.67
No / not reported 57 0.54 (0.45-0.63) 25 (0~46)
TE 29 0.55 (0.43-0.67) 50 (16~67) 5 1.08 (0.77-1.40) 73 (0~87)
PUBLIATION BIAS
Adjustment for Add 27 Egger’s regression Add 21 Egger’s regression Add 13 Egger’s regression
EP
4
publication bias 0.41 (0.33 – 0.50) intercept p < 0.01 0.49 (0.37-0.61) intercept p = 0.01 0.71 (0.54-0.89) intercept p < 0.01
1 2 3
selfhelp treatments excluded; studies not reporting this were excluded (WASO, n=2; SQ, n=2); studies not reporting inclusion range for age were excluded (WASO, n=6; SQ,
4 5
n=3); trim and fill procedure which estimates the number of missing studies (number that needs to be added) and produces the effect size after adding those studies; the 95%
C
2
CI of I cannot be calculated when the number of groups is lower than three; * p < 0.05, ** p < 0.01; F2F = face-to-face
AC
2
ACCEPTED MANUSCRIPT
Table 4. Association between post-test treatment effects (SOL, SE and ISI) and treatment-, patient- and study variables (standardized regression
coefficients of multivariate metaregression analyses)
Sleep Onset Latency (SOL) Sleep Efficiency (SE) Insomnia Severity Index (ISI)
PT
g (95% CI) P g (95% CI) P g (95% CI) p
TREATMENT VARIABLES
Full CBTI
Yes Ref 0.56 n/a n/a
RI
No 0.08 (-0.18 -0.33)
Treatment format
F2F individual Ref 0.62 Ref 0.03 Ref 0.05
SC
F2F group -0.04 (-0.23-0.16) 0.33 (0.04-0.62) -0.47 (-1.01-0.08)
Selfhelp -0.13 (-0.38-0.13) -0.05 (-0.36-0.26) -0.56 (-1.01- -0.11)
PATIENT VARIABLES
U
Population
Community Ref 0.55 Ref 0.28 n/a
AN
Other 0.06 (-0.14-0.26) 0.16 (-0.13-0.45)
Comorbidity
Yes n/a Ref 0.78 n/a
No -0.05 (-0.36-0.27)
M
Sleep medication
Allowed Ref 0.28 Ref 0.65 Ref 0.34
Not allowed 0.02 (-0.18-0.23) 0.13 (-0.15-0.41) -0.37 (-0.87-0.14)
D
Not reported 0.28 (-0.07-0.63) 0.06 (-0.56-0.69) -0.01 (-0.94-0.92)
Age TE
Younger adults only Ref 0.42 Ref 0.01 Ref 0.02
All adults -0.17 (-0.41-0.08) -0.28 (-0.62-0.06) -0.61 (-1.07- -0.14)
Older adults only -0.20 (-0.46-0.06) -0.22 (-0.58-0.14) -0.28 (-1.00-0.43)
EP
Not reported -0.14 (-0.38-0.11) -0.70 (-1.13- -0.27) -0.81 (-1.43- -0.19)
STUDY VARIABLES
Year of publication -0.001 (-0.012-0.009) 0.85 0.02 (-0.001-0.05) 0.06 n/a
Type of control
C
1
ACCEPTED MANUSCRIPT
Allocation concealed
Yes Ref < 0.01 n/a Ref 0.02
Not reported 0.27 (0.08-0.46) 0.45 (0.09-0.81)
Study drop-out
0-9% n/a Ref 0.80 Ref 0.35
PT
10-19% 0.09 (-0.16-0.35) 0.22 (-0.15-0.59)
20% or more 0.13 (-0.17-0.42) 0.20 (-0.23-0.63)
unknown -0.11 (-0.90-0.67) -
RI
ITT analysis n/a
Yes Ref 0.61 n/a
No/not reported -0.07 (-0.33-0.19)
SC
R squared 0.30 0.32
95% CI = 95% confidence interval; CBTI = cognitive behavioral treatment; F2F = face-to-face; g = hedges g; ITT = intention to treat
U
AN
M
D
TE
C EP
AC
2
ACCEPTED MANUSCRIPT
PT
RI
Removing duplicates
(n = 290)
SC
Records screened (n = 1.727) Records excluded (n = 1.503 )
U
AN
Full-text articles assessed for
eligibility (n = 224)
Full-text articles excluded (n = 137 )
• No RCT (n=21)
• No CBTI components in therapy (n=23)
• No waitlist/care-as-usual comparison (n=28)
M
• Secondary analysis (n=20)
• Tapering medication (n=5)
Studies included in meta-analysis • Other (n=40)
D
(n = 87)
TE
EP
C
AC